MasterClass on ICT Support for Adaptivenessand (Cyber)Security in the Smart Grid 

DAT300, DIT668

Introduction: The “Smart” Grid as Distributed 

Cyberphysical system& Course Outline Marina  Papatriantafilou Networks and Systems Division

Computer Science and Engineering DepartmentChalmers University of Technology & Gothenburg University

Application domains: energy systems, vehicular systems, communication systems and networks 

Briefly on research + education area of the supportingteam

Distributedproblems  over network‐based

systems(e.g. overlays, 

distributed, locality‐based resourcemanagement) 

Reliability, adaptiveness, 

securitySurvive failures, detect  & mitigate attacks, secure self‐organization, …

ParallelprocessingFor efficiency, 

data&computation‐intensive systems, programming new 

systems(e.g. multicores)

Magnus Almgren M. PapatriantafilouOlaf LandsiedelVincenzo Gulisano

Viktor Botev Valentin Tudor 

A Distributed SystemA set of computing&communicating processes, 

collaborating for acheiving local and/or global goals

3

A Distributed System

Figure: Computer Networking: A Top Down Approach , Jim Kurose, Keith Ross, Addison‐Wesley.4

A Distributed System

Picture sourec:Wikipedia

5

A Distributed System

6

A Distributed System

Pic. http://www.hatswitch.org/~sn275/courses/DSS/info.shtml7

A Distributed Cyberphysical System

Picture source http://www.energy‐daily.com/images/smart‐grid‐electricity‐schematic‐bg.jpg8

the story: in the traditional El Grid…

Dispatchable Sources Oblivious Loads

Generation

Transmission Distribution

….while in the greener El Grid

Dispatchable Sources

Generation

Transmission Distribution

+ aware, Interactive loads+ non‐dispatchable, distributed   sources

+ info, communication, computation+ emeters

• paradigm shift: – from pre‐planned broadcasting, to adaptive scheduling

Cyberphysical systems as layered systems

Cyber system 

El- link and/or communication link

Overlay network

Physical system 

Computing+ communicating device

CPS => enormous numbers of devices & analogousvolumes of data!

11

Why this complexity? 

Wise/adaptive use of resources ….… possibilities of optimization: e.g. energy consumption, security 

services,… increasing the quality of life” [the 4th industrial (r)evolution, presentation S. Jeschke, 2013]

… or  ”smart” cities …. 

… or ”some V’s … 

• Volume: terabytes – peta/exa/zetabytes

• Velocity: streams

• Variety: various types of data …

… system: Big!   … data: Big!  (but: locality!)

… and one D”: DistributionNot always necessary to centralize => allow multiple actors, data‐streaming, scaling, 

privacy, …

i.e. BIG!

Good! Process on‐the fly can eg filter peta+bytes to megabytes

with various relevance domains; good!

Cyberphysical systems:possibilities and challenges shake hands

Cyber system 

El- link and/or communication link

Overlay network

Physical system 

Computing+ communicating device

A bit of info on the el‐network

15

The traditional Electrical Grid

Generation Transmission Distribution

Managed and monitored by the SCADA system. No dedicated real time monitoring system (yet). Pic. V. Tudor

16

From centralized to distributed generation

Power IslandPic. V. Tudor

17

One of the enabling components: Smart Meter (Advanced Metering Infrastructure)

A “Smart” Meter: is a small embedded system automates (consumption) index

readings instantaneous consumption in-door display time of use tariffs the base for the AdvancedMetering Infrastructure

Pic. V. Tudor

18

Smart Meter components

Pic. V. Tudor

In the Power Grid cyber‐layer

Selected topics:

• Adaptiveness: Distributed resource management

• Enabling “tools”: Communication, Data, information 

• Orthogonal and utterly important: cyber‐security

19

In the Power Grid cyber‐layer

Selected topics:

• Adaptiveness: Distributed resource management– Demand‐side management: load balancing, load shifting (users)

– Routing, aggregation (network)

• Communication, Data, Information 

• Cybersecurity

20

Demand-side managementhousehold/neighborhood-scale and more

……

…

……

…

21

Problem: Fine‐grained align supply & consumption; continuous decisions based on info on load, availability, constraints, possibilities ((non)shiftable load, thermal or other storage…)

Vision for microgridsfor better use of renewables: Virtual Private Grids/microgrids• communicating supplies and loads • cooperating for 0‐net energy or 

mixed use of renewable and other sources 

• adaptive loads, to draw power when renewables provide it

• ie connect to the aforementioned methods are for, plus– Power routing ([NKGPLB10] and  

aggregation– Information!....

Picture sources: http://www2.ee.ic.ac.uk; Katz etal Sustainable computing 2011

22

In the Power Grid cyber‐layer

Selected topics:

• Distributed resource management

• Enabling “tools”: Communication, data, information – Distributed sources & processing – Wireless/sensor networks– Monitoring, facilitating resource services

• Cybersecurity23

DataBase

ElectricCar

SmartMeter

Collector

Information&Communication, Advanced Metering Infrastructure

Local (info aggregation) ”agents” can • Make sesnse out of data, Monitor 

(DATA AGGREG)• influence power/energy 

distribution (POWER AGGREGATION in collab w. DATA AGGR)

24

Data gathering/processing  in Sensor Networks

nodes produce relevant information about their vicinity periodically.

Data is conveyed to an information sink for further processing.

….Routing

On which path is node u’s data forwarded to the sink?

Processing/streaming/aggregation

… data can be processed as it is routed to the collector/aggregator (sink).

In-network aggregation/streaming/processingWhere/how is u’s data processed?

Work with routing, streaming, coding, processing schemes to deliver needed info to the sink (care also for privacy).

DataBase

ElectricCar

SmartMeter

Collector

Information/Communication

• Distributed methods for data‐network reliability

• Aggregate distributedly, at different hierarchy levels, for detecting situations, with:• Scalability• Parallelism for massive data

27

In the Power Grid cyber‐layer

Selected topics:

• Distributed resource management

• Communication, information 

• Orthogonal issues: cyber‐security• Extra important for overall system reliability

28

Imperative to addresscyber security from the start! [F10]

Back Office

ElectricCar

AppliancesAlarmSystem

SmartMeter

Collector

Accidents/unwanted situations attempts arepossible

Lesson learned from Internet

29

Distributed system/Collaborative Security methods can help prevent/detect/ mitigate malicious attempts

Cybersecurity aspects  Case studies

Possible to destabilize parts of the system by inappropriate access to e.g. remote on/off possibilities [TKAPS11]

Avoid the Internet examples of defacto standards info‐security from the start Distributed/collaborative security methods can help to deal with scale 

30

One of the ideas/hypothesis: cyber security and distributed  information processing

Back Office

ElectricCar

AppliancesAlarmSystem

SmartMeter

Collector

Unwanted/malicious

attempts arepossible

• Recall aggregation at different levels + parallelism:  

• Can also help for privacy,localization&mitigationof unwanted situations. 

31

Strategic relevance for research and education? 

“… area of strategic importance … advances fast,  technologically& commercially…

Cisco expects the Smart Grid communication network will be 100 eller 1000 times larger than the Internet”

[Vinnova, ”SmartaNät” 2011 ]and references therein

“… any vulnerability within this software‐intensive critical system will attract attention from hostile groups …” 

[MSB, “If one goes down all goes down?”, 2010]

32

• large investments• off‐the‐shelf  info/software solutions are not there

• careful, informed, multidisciplinary expertise needed in deployment

• cf. lessons learned from Internet 

• Distributed computing and systems  and Security in the core of the cyberphysical infrastructure

Strategic relevance for research and education? 

33

Our related research projects in a nutshell

Advanced Metering Infrastructure‐Metropolitan‐scalenetworks‐Data validation‐Extracting info from the data,  detect unwantedsituations‐Security: strengthen, work with encrypted data

Demand‐responseResourcemanagement, loadshaping, with/outstorage&forecastsWesternHarbor/HyllieInfo/data to enablepossibilities

Energy/efficientcomputationimprovements in energy efficiency for computing systems:HPC‐centers &embedded devicesest savings 30‐70%%

Computation/big data

Projects’ support & collaboration

Course/Masterclass:ICT Support for Adaptiveness and Security in the 

Smart Grid (DAT300)• Goals

– students (from computer science and other disciplines) get introduced to advanced interdisciplinary concepts related to the smart grid, thus 

– investigating a domain‐specific problem relevant to the smart grid that need an understanding beyond their traditional fields.

Idea

• Based on both the present and future design of the smart grid. – How can techniques from distributed systems be applied to large, heterogeneous systems where amassive amount of data will be collected?

– How can such a system, containing legacy components with no security primitives, be madesecure when the communication is added by interconnecting the systems?

• The students will have access to a hands‐on lab, where they can run and test their design and code.

In this course: Topics: • Adaptiveness: Distributed resource management• Enabling “tools”: Communication, Data, Information 

processing• Cyber‐security

Structure, todo’s:• Projects • Guest lectures by the supporting team + industry and 

related parties• Self‐study and presentations

How? • Cf Administrative Details.pptx

38